cloud computing data center savageau1

7/29/2019 Cloud Computing Data Center Savageau1

1/27

Pacific-Tier Communications

Data Center ConsolidationDisaster Recovery

Cloud Computing

Discussion and Overview

September 9, 2010


John R. Savageau

President


2/27


Discussion Topics

Data Center Consolidation

Power

Facility Planning

Facility Requirements Data Center Infrastructure Plan

Primary processing sites

Disaster recovery

Move to Virtualization and CloudTechnologies

Develop a national data center consolidation and disaster recovery plan


3/27


Government Data Center Reviews

Individual ministries and agenciesare operating critical nationalsystems and applications in smallcomputer rooms and server closets

Data is either backed up locally (notoffsite) or not backed up at all

Limited or no redundancy forapplications, power and telecom

Pacific Tier Interviews in Vietnam, Indonesia, Moldova, Hawaii, Mongolia


4/27


Government

Data Center

ConsolidationNot just an issue for

Developing

countries it is a

common issue for

most governments


5/27


Government Visions & Trends

The starting point is this: in central government wehave 130 data centers. Following a survey of central

governments insourced and outsourced data

centers, we asked the question: Do you need 130?The analysis was done by IT industry trade bodyIntellect. It said that we could significantly reducethe number from 130 to something like 10-12.

(John Suffolk/UK)


6/27


Data Center Consolidation

Adopt modern technologies and practices that willimprove the effectiveness and efficiency of datacenter use

Consolidate data centers in a model that willreduce duplication and un-necessary costs forbase ICT infrastructure

Match national objectives and business needs toICT utility and protection of national data

VISION: Create a whole-of-government approach to data centers,

disaster recovery, and business continuity. Provides standards andintegration of the governments ICT infrastructure matching a wide range

of government information needs in an optimal manner with regard to

interoperability, development agility, cost, energy use, and operations.

Australia


7/27Pacific-Tier Communications

Consolidation Initiatives

Consolidation Type Objectives Cost Savings

Physical Consolidation Reduce number of data centers throughnational data center colocation Migrate equipment to national data

center locations with more favorable

operating costs

Facility costs Hardware and Software maintenance

Non-application specific admin staff

Reduced unit costs

Equipment Consolidation Replace existing hardware systems witha national standard

Hardware costs

Hardware maintenance

Reduced unit costs

Server/Storage Optimize the number of existinghardware platforms

Replace low-end and inefficient servers

with higher performance and cheaper

systems

Share backup, mirroring, and disaster

recovery platforms

Number of servers

Number of storage and backup devices

Hardware maintenance

Storage systems

Reduced unit costs

Applications Library Standardize basic applications usedacross government Software maintenance Software licensing Immediate availability

Communications Provide multiple carriers and serviceproviders to interconnect facilities/users

High level of communications

redundancy to provide 100% availability

Promotes price competition



Sun Consolidation Example

Consolidated multiple European data centers intoa single, UK facility Achieved an 80% operational space reduction

50% reduction in electrical power and cooling costs

On a worldwide level the move to new technologyhas enabled Sun to: Reduce 267,000 square feet of data centre to 133,000 sqft

Save around 4,100 tonnes of CO2 per year

Consolidated design Cut power costs by 66% Increased processing power by around 450% (w/virtualization)

Intellect/UK

The Economics of DC Consolidation



Consolidation Supports Risk Management

Centralized risk management

Continuous monitoring (including security)

Immediate nation-wide response to disasters Cost savings

Single point of contact for vendor support

Rapid expansion and technology agility

Standardization and Interoperability

US-CIO



Why Some Resist Consolidation

NASCIO



Data Center Standards

Standards differ primarilyin terminology

Recognized as thestandard data center

architecture for mostenterprise

Universally understood byconstruction,

commissioning, andsecurity

Promotes standardization

Computer Room

Data Center Floor

Network and Building

Operations Center

Telecom Room

(Main Distribution

Frame) Serving Data

Center Spaces

Storage Rooms and

Loading Docks

Support Staff Offices Entrance Room/s

Data Center

Electrical and

Mechanical Rooms

General Office Space

Telecommunications

& Equipment Rooms

Serving Spaces

Outside of the Data

Center

Building Site

Building Shell

Data Center

TIA-942 Relationship of Spaces within a Data Center

ISO/IEC 24764, TIA-942, BICSI 002-2010



Disaster Recovery



Disaster Recovery Planning

International DR/DRC/BCP standards NFPA 1600 (Process and Planning)

ISO 24762

BS 25999 (UK)

COBIT/ITIL

Main objectives

Safeguard data

Respond quickly

Restore operations

- Monitor

- React

- Restore- Recover



Tiered Disaster Recovery Model

Tier 1: Data backup with no hot site Off-site storage facility

Manual process (tapes)

Tier 2: Data backup with a hot site Backups on tape

Restore systems from backup tapes in the event of a disaster

Recovery time is more predictable

Tier 3: Electronic vaulting Mission critical data is electronically vaulted

High-speed communication circuits connect to an automated tape library at a remote site

Tier 4: Point-in-time copies Disk-based backup

Higher frequency network backups to disks

Tier 5: Transaction integrity Consistency of data between the production and recovery data centers

Little to no data loss in such solutions

Dependent on the application in use.

Tier 6: Zero or near-zero data loss Highest levels of data currency disk mirroring

Tier 7: Highly automated, business integrated solution Integration of automation

Recovery of the applications is automated

No loss in application performance for end users

Recommendedfor critical government systems and applications

Tier Zero

No business continuity plan.

There is no saved information,

no documentation, no backup

hardware, and no contingency

plan. The time necessary torecover in this instance is

unpredictable. In fact, it may not

be possible to recover at all.

This applies to several ministry-

level data centers visited over

the past year.


15/27


Best DR/BCP Plan for Developing Nations

Primary

Backup

Archive

Primary processing location

Backup processing location Mirrors primary processing location

Can be used for load balancing

Remote storage and archival Data files

SaaS library images

Allows government

operations continuity in theevent of major disruption


16/27


Cloud PlanningVirtualization and Cloud Computing

Strategies for Developing Nation

eGovernment Plans


17/27


US Government View on Cloud Computingas the worlds largest IT user

If an individual can create a free email account

in a matter of minutes, and a small business can

create its entire financial system online in acouple minutes, then why must the government

spend billions of dollars building (similar)

systems that may not be sensitive in nature?

Vivek Kundra, the US Governments Federal Chief Information Officer


18/27


Issues and Concerns w/Clouds

Security of government information Personal Privacy of citizens

Availability and performance

Management and control Migration planning

Cloudcomputing is not a silver bullet,

but offers a transformational opportunityto fundamentally reshape the operations

of government and close the ITgap(Kundra)


19/27


Cloud Characteristics

On-demand and self-service

Broad network access

Resource pooling Rapid elasticity

Measured service

From NISTs Cloud Computing Definition

Refresher


20/27


What a Cloud Brings to Governments

A manageable pool of computeand storage resource (IaaS)

An environment allowing creation

of virtual data centers (PaaS) Standardization and rapid

provisioning of applications

(SaaS) Greater control over security

Greater control over OPEX/CAPEX


21/27


Government Benefits of SaaS

Data is stored on protected media

Reduce the cost of IT workstations andequipment

Reduce the risk of viruses and malware

Central management of software versionsand updates

Application and data standardization

Enhance potential for systems integrationand data transparency across agencies

Shared network and telecom resources


22/27


Government Roles in Cloud Create national policies and standards

Data security

Data formats

Data and IT Governance

Produce cloud computing solutions insupport of government operations

Provide guidance for both government andcommercial initiatives

Strategic roadmaps

Best practices

PPPs


23/27


Moldova Issues

Obsolete or unsupported h/w, s/w

Incompatible systems

Old or non-existent documentation

Loss of skilled or experienced staff Inability to meet business requirements

Data security shortfalls

Disaster recovery

Standard access controls


24/27


Potential Cloud Plan for Moldova Distributed architecture

Cloud processing mirrored in both primary and backup data centers Both locations are real time

Disaster recovery is built into the distributed architecture

Processing on demand for ministries and othergovernment agencies Virtual data centers Storage and disaster recovery support

Standard applications library Provides standardized applications throughout national ICT infrastructure

Cheaper, maintainable, and promotes data mining for central government

Promotes hosted and web-based applications Pushes storage and files into data center, away from laptops and

unsecure media

Transparency


25/27


The Road to Cloud Moldova

Short term take away risk to data and facilities

through central backup/DR and consolidation Cloud steps

Designate and empower CIO/ICT leadership Project management

Operations management (could be CTS)

Develop project policy and priorities

Identify pilot projects

Process and integration engineering

Technology and standards development

Develop standardized applications SaaS

Database Where possible, re-engineer withCommercial-off-the-Shelf

(COTS) software/applicationsor supported Open Source


26/27


Cloud High Level Applications

Relieves the end-userfrom the burden ofinfrastructure

Focus CAPEX/OPEX oncore business

National

Data Center

Shared

Telecom

Infrastructure


27/27


Questions?John R. Savageau

[email protected]

+1-808-271-1372
mailto:[email protected]:[email protected]:[email protected]:[email protected]

cloud computing data center savageau1

Documents